Intensification of ATRP polymer syntheses by microreaction technologies

(1)

HAL Id: tel-01124285

https://tel.archives-ouvertes.fr/tel-01124285

Submitted on 6 Mar 2015

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Intensification of ATRP polymer syntheses by

microreaction technologies

Dambarudhar Parida

To cite this version:

(2)

Thèse présentée pour obtenir le grade de Docteur de l’Université de Strasbourg

Discipline : Chimie des matériaux

Spécialité : Génie des procédés de polymérisation Ecole doctorale : Physique et Chimie – Physique

Présentée par:

Dambarudhar PARIDA

__________________________________________________________________________________

Intensification of ATRP polymer syntheses by microreaction technologies

__________________________________________________________________________________

Soutenue publiquement le 13 février 2014

Directeur de Thèse : C. A. SERRA Professeur, Université de Strasbourg Rapporteur Externe : T. F. L. MCKENNA Directeur de Recherche, CNRS - CPE Lyon Rapporteur Externe : A. RENKEN Professeur, EPFL Lausanne

Examinateur Interne : J.-F. LUTZ Directeur de Recherche, CNRS - ICS Strasbourg

(3)

Acknowledgements

Since we started working as a team and exchanging ideas, we are moving at much greater pace than ever before. Sometimes, we travel far in search of ideas or to share our ideas and it is happening for centuries. Following such a route, I came to Strasbourg to sharpen my knowledge and skill. This simple looking route was never easy without a team. They worked like a support, when I was weak and they helped me to grow stronger. Their guidance helped me in my research endeavour. Along with human factors there was a need for financial support to carryout research for which ANR grant n° 09-CP2D-DIP² is greatly appreciated. I would like to thank Dr. T. F. L. Mckenna, Prof. A. Renken, Dr. J.-F Lutz for accepting to become a member of jury.

Prof. C.A. Serra at ECPM, University of Strasbourg, thank you for being amazing guide/Guru and giving me an opportunity to explore the field of microfludic and controlled radical polymerization. You have been a continuous source of inspiration and motivation for me. Ever since I arrived at Strasbourg, You have supported me not only in research but also emotionally through the ups and downs to finish the thesis, especially towards the end. Every time I came with a proposal, always I got the answer “You have my green light”. Thank you for moral support and freedom I needed to move on. Special thanks to Isabelle for her support and encouragement.

Dr Florence. Bally, Université de Haute Alsace, Thank you for introducing me to micro and CORSEMP. You always welcomed my queries and shared your experiences which helped me to overcome small yet significant obstacles. Truly, it was agreat experience working with you. I would like to thank Dr. Yannick Hoarau and Dhiraj K. Garg for their observations and reasoning made while working together which helped me to answer some important questions. Dhiraj, working with you was really fun and exciting.

(4)

free travel to different conferences was possible because of your effort. Merci Cheng and Catherine for your continuous support.

Dozens of people taught me and helped me in many ways at ICPEES. Dr. Anne was always ready to listen and discuss the problems. You helped me in many ways to think from a different perspective. Dr. Nicolas, your tracer was really helpul; you were like a ray of hope during that time. I cannot imagine determining residence time distribution without the tracer. Prof. Muller, Dr. Rigoberto and Dr. M. Bouquey shared their experiences and expertise with me for rheological analysis of samples. It helped me to have a look in side the microreactors. Discussion with Dr. Rigoberto on world religion was also very interesting. Special acknowledgement for Prof. Muller, Dr. Nicolas, Dr, Anne, Dr. Rigoberto and Dr. M. Bouquey.

In the lab there were many amzing people around me who taught many things in many ways. Mumu is a great friend ever since she came to our lab and has been supportive in every way. Yu Wei, a nice co-worker who like to talk and solve different problems of the lab. It was interesting to discuss chineese philosophy and lifestyle with him. Salima, a smiling, helpfull and talented friend. Thank you for bringing bollywood to Strasbourg. FX; Alexandar, Dr. Mathieu, Dr. Patricia, Dr Korine, Ibrahim, Lucas...thank you for all your support for all these years.

People in the office R1 H2 B1 are not mentioned yet, because they deserve their own part. I learned a lot from them. They were always there to help me in every possible and impossible ways. They were always ready to listen and discuss on various topics and issues, be it polymer or politics. Our discussions on polymers, helped me to shape my experiments in a significant way. I will remember 4:00 pm as the time for Technologic or Philosophy or Cricket. I realy appreciate their efforts for tolerating me in the office and teaching me French. Thank you Dr. TGV for gâteau au chocolat (Monday Special), Alice for gâteau aux pommes (Though you made it Twice) and Ikhram for biscuits. Marie, you came late to the office but soon you became a good friend. Stephanie, Alice, Ikhram, Marie, Mumu...you people registered my stay at Strasbourg in the unforgettable part of my memory. Even hundred pages are not sufficient to describe what you did for me.

(5)

(6)

French summary

Résumé de Thèse

Intensification de la synthèse de polymères par ATRP au moyen de

technologies de microréaction

1. Introduction

Contrairement à tout autre composé chimique, les macromolécules synthétiques ont des caractéristiques (masse molaire moyenne, distribution des longueurs de chaîne) qui sont fortement dépendantes des paramètres du procédé de synthèse. L’augmentation de la viscosité, qui peut atteindre jusqu'à 7 décades pour les procédés en masse et en solution concentrée, s’accompagne d'une diminution des transferts de matière et de chaleur. Cette situation conduit généralement à l'élargissement de la distribution des longueurs de chaînes, une masse molaire différente de celle souhaitée et peut également entraîner l’emballement thermique du réacteur. Cependant, un mélange rapide des réactifs en entrée de réacteur et en son sein est souvent conseillé pour résoudre ces problèmes.

Tant du point de vue du mélange que de l’évacuation des calories libérées par la réaction, les micromélangeurs et microréacteurs peuvent être une option pour surmonter les limitations diffusionnelles. En effet, la très faible dimension caractéristique (10 à quelques centaines de micromètres) de ces systèmes microfluidiques leur confère un avantage certain par rapport à leurs homologues en verre de laboratoire et leurs versions industrielles. A cette échelle, il a été constaté que les microsystèmes peuvent améliorer considérablement les transferts de masse et de chaleur. Comparés à d'autres produits chimiques, la synthèse de polymères dans les systèmes microfluidiques est relativement nouvelle et peut largement bénéficier des caractéristiques précitées de ces microsystèmes (cf. chapitre 1).

(7)

French summary

méthacrylate de 2-(diméthylamino)éthyle (DMAEMA) servira de cas d’étude tout au long de ce travail de thèse. L’objectif de cette dernière étant d’intensifier la production de (co)polymères linéaires et branchés au moyen de systèmes microfluidiques et des paramètres de procédés.

2. Effet d’un prémélange sur la polymérisation en microréacteur hélicoïdal

Afin d’étudier l’effet de différents principes de micromélange sur les caractéristiques de copolymères statistiques, plusieurs types de micromélangeurs (à bilaminaion, jonction en T ; à jet d'impact, KM Mixer ; à multilamination interdigitale, HPIMM) ont été considérés pour la copolymérisation par ATRP du DMAEMA et du méthacrylate de benzyle (BzMA) de compositions différentes (20 et 40 mol.% BzMA) dans un microréacteur hélicoïdal (CT, schéma 1.a) de diamètre interne 876 µm (Figure 1). D'après les résultats obtenus, l'impact du prémélange est évident puisque, toute chose égale par ailleurs, les copolymères ont des propriétés différentes en fonction du type de micromélangeurs employé. La jonction en T a conduit à la conversion en monomère la plus faible alors que le micromélageur à multilamination a donné les conversions les élevées (+ 30% par rapport à un réacteur discontinu) tout en assurant des masse molaires plus élevées et une distribution des longueurs de chaînes plus étroite. Les détails sont expliqués dans le chapitre 3.

(a) (b)

(8)

French summary 0 5000 10000 15000 20000 25000 0 50 100 150 200 250 M n ( g /m o l) Time (min.) Batch 20% Batch 40 % T-J 20% T-J 40% HPIMM 20% HPIMM 40% KM 20% KM 40%

Figure 1. Evolution de la masse molaire moyenne en nombre des copolymères

synthétisés pour différents temps de séjour, micromélangeurs et compositions molaire en BzMA.

3. Accélération de la cinétique ATRP en microréacteur hélicoïdal

(9)

French summary 1.3 1.5 1.7 1.9 2.1 0 5000 10000 15000 20000 25000 0 20 40 60 80 P D I M n ( g /m o l) Conversion (%) 60 °C 75 °C 85 °C 95 °C

Figure 2. Evolution de la masse molaire moyenne en nombre (symboles pleins) et du PDI

(symboles vides) avec la conversion du monomère pour différentes températures de polymérisation.

(10)

French summary 1.2 1.3 1.4 1.5 1.6 1.7 1.8 0 5000 10000 15000 20000 25000 0 25 50 75 100 P D I M n ( g /m o l) Pressure (bars) 576 µ 876 µ 1753 µ

Figure 3. Effet de la pression sur la masse molaire moyenne en nombre (symboles pleins) et

le PDI (symboles vide) pour différents diamètres de microéacteur.

(11)

French summary 0 10 20 30 40 50 60 70 80 0 25 50 75 100 125 C o n v e rs io n ( % ) Time (minute) 3 mtr. 6 mtr. 9 mtr. 18 mtr.

Figure 4. Effet de la longueur du microréacteur et du temps de séjour sur la conversion du

DMAEMA.

Toutes ces observations sont détaillées dans le chapitre 4.

4. Influence de la géométrie du microréacteur

(12)

French summary 1 1.2 1.4 1.6 1.8 2 0 5000 10000 15000 20000 25000 0 20 40 60 80 P D I M n ( g /m o l) Conversion (%) CT CFI

(symboles vides) avec la conversion du monomère pour différents microréacteurs tubulaire.

Le bénéfice de la technique d’inversion de flux est encore plus marqué lorsqu’on étudie la synthèse d’un polymère branché. Il a ainsi été observé, lors de l’incorporation dans le mélange réactif initial d’un inimère (molécule capable d’agir comme un monomère et un amorceur), que la masse molaire obtenu augmentait et que le PDI diminuait très significativement (figure 6). Par ailleurs, le taux de branchement, déterminé par chromatographie d’exclusion stérique, était sensiblement plus élevé dans le cas de l’emploi du CFI que pour le CT ou bien le réacteur discontinu et cela pour des taux d’inimère allant jusqu’à 10% en mole. 1.5 2 2.5 3 1000 1500 2000 2500 3000 3500 0 20 40 60 80 100 120 P D I M n ( g /m o l) Time (min.) Batch CT CFI

(13)

French summary

Afin d'améliorer la productivité horaire du procédé de synthèse du PDMAEMA par ATRP dans un réacteur tubulaire, de plus grands diamètres internes ont été testés (1753 et 4084 mm) à temps de passage constants. L’augmentation du diamètre engendra certes une augmentation de la quantité horaire de polymère synthétisé mais s’accompagna d’une diminution marginale de la conversion du monomère (figure 7) et surtout d’une augmentation importante du PDI. Pour limiter cet effet néfaste, nous avons eu recouru à la technique d’inversion ; et pour bénéficier au mieux du mélange interne promu par cette technique, nous avons également augmenté la longueur du réacteur. Ainsi ces plus « gros » CFI ont engendré non seulement un accroissement du taux de conversion du monomère (figure 7) et de la masse molaire mais surtout une forte réduction du PDI et cela au prix d’une augmentation quasi négligeable de la perte de charge, c’est-à-dire de l’énergie requise.

15 35 55 75 95 0 30 60 90 120 C o n v e rs io n ( % ) Time (min.) CT 876 µm CT 1753 µm CT 4083 µm CFI 876 µm CFI1753 µm CFI 4083 µm

Figure 7. Evolution de la conversion du monomère en fonction du temps pour différents types

de réacteurs et diamètres et une longueur de 3m.

Le chapitre 5 explique toutes ces observations dans le détail.

5. Conclusion

La mise en œuvre de synthèses de (co)polymères par la technique ATRP dans des systèmes microfluidiques a permis de mettre en évidence les aspects suivants :

(14)

French summary

- les paramètres du procédé comme la température ou la pression peuvent accélérer de manière significative la réaction de polymérisation. Une fenêtre étroite existe pour laquelle l’augmentation de température présente un effet bénéfique. Une augmentation modérée de la pression (100 bars) engendre un contrôle accru sur la polymérisation. L’effet du cisaillement lors de la polymérisation s'est révélé être dépendant de la longueur de chaîne. L'augmentation de la longueur de chaîne a tendance à réduire l'effet bénéfique du cisaillement. Ainsi l’augmentation du cisaillement d’un facteur 6 a peu d'effet sur la conversion du monomère au bout de 2 heures de temps de polymérisation mais l'effet est plus prononcé pour de faibles masses molaires (temps de séjour inférieurs à 1 heure).

- une très nette amélioration du contrôle de la polymérisation a été observée avec un changement de la géométrie du microréacteur tubulaire. L’introduction d’inversions de flux à intervalles réguliers le long du microréacteur a ainsi permis de réduire sensiblement le PDI. La polymérisation en CFI a également montré une augmentation significative de l'efficacité de branchement qui est un indicateur de l'amélioration de l'architecture branchée. Enfin l’emploi de CFI de plus grand diamètres a validé la possibilité d’augmenter la productivité horaire au détriment cependant d’une petite perte de contrôle sur les caractéristiques macromoléculaires, toutefois moins importante que dans le cas des CT.

(15)

(16)

Abbreviations and notations

Abbreviations

1_{H NMR} _{Proton nuclear magnetic resonance}

ATRP Atom Transfer Radical Polymerization

BIEM 2-(2-bromoisobutyryloxy)ethyl methacrylate

BzMA Benzyl methacrylate

CFI Coil flow inverter

CORSEMP Continuous online rapid size-exclusion monitoring of polymerization

CRP Controlled radical polymerization

CT Coiled tube

CuBr Copper (I) bromide

DCM Dichloromethane

DMAEMA 2–(Dimethylamino) ethyl methacrylate

DMF Dimethylformamide

DP Degree of polymerization

EBIB Ethyl 2–bromoisobutyrate

FRP Free radical polymerization

GPC Gel Permeation Chromatography

HEMA 2-hydroxyethyl methacrylate

HMTETA 1,1,4,7,10,10–hexamethyltriethylenetetramine

HPIMM High Pressure Interdigital Multilamination

Micromixer,

HPLC High performance liquid chromatography

KM Impact jet micromixer

MALS Multi angle light scattering

NMP Nitroxide mediated polymerization

PDI Polydispersity index

PDMAEMA Poly(2–(dimethylamino) ethyl methacrylate)

PMMA Poly(methyl methacrylate)

RAFT Reversible Addition-Fragmentation chain Transfer

(17)

Abbreviations and notations

RI Refractive index

RTD Residence time distribution

SCVCP Self condensing vinyl copolymerization

SI-ATRP Surface Initiated Atom Transfer Radical

Polymerization TEA Triethylamine THF Tetrahydrofuran

Notations

C Polymer concentration (g/L) C* _{Overlap concentration (g/L)}

KATRP Equilibrium constant (-)

kp Propagation constant (mol/L/s)

kt Termination constant (mol/L/s)

L Tubular reactor length (m)

Mn Number-average molecular weight (g/mol)

Mw Weight-average molecular weight (g/mol)

Mw MALS Weight-average molecular weight as seen by multi

angle light scattering detector (g/mol)

Mw RI Weight-average molecular weight as seen by refractive

index detector (g/mol)

n Flow index of power-law fluids (-)

Q Volume flow rate (m3_/s)

R Tubular reactor radius (m)

η Polymer solution viscosity (Pa.s)

η0 Solvent viscosity (Pa.s)

ηsp Specific viscosity (-)

[η] Intrinsic viscosity (L/g)

W

(18)

(19)

1. INTRODUCTION

2

2. BACKGROUND LITERATURE

Preface 6

2.1 Aspects of controlled radical polymerization (CRP)

7

2.1.1 General overview 7

2.1.2 Special considerations for ATRP 8

2.1.3 Intensification of CRP in batch reactors 9

2.1.3.1. Acceleration of NMP 9 2.1.3.2. Acceleration of RAFT 10 2.1.2.3 Acceleration of ATRP 12 2.1.4 Summary 14

2.2 Microreaction technology

14 2.2.1 Microreactors 16 2.2.1.1 Different types 16 2.2.1.2 Special features 19

2.2.2 Microdevices for polymer synthesis 23

2.2.2.1 Microreactor setup 23

2.2.2.2 Micromixers 25

2.2.3 Benefits of microdevices for polymers and copolymers synthesis 31

2.2.3.1 Polymers with controlled macromolecular characteristics 32

2.2.3.2 Polymers with controlled chemical composition 40

2.2.3.3 Polymers with controlled architecture 45

2.2.3.4 New operating windows 49

2.2.4 Online monitoring 55

2.2.5 Summary 55

2.3 Conclusion

56

References

57

3. MATERIALS AND METHODS

Preface 68

3.1 Materials

69

3.1.1 Main reagents and chemicals 69

3.1.2 Synthesis of 2-(2-bromoisobutyryloxy)ethyl methacrylate 69

3.2 ATRP in batch reactor

70

3.2.1 Linear homo polymerization 70

(20)

3.2.3 Branched PDMAEMA synthesis 71

3.3 ATRP in microreactor

72

3.3.1 Microreactor setup and instrumentation 72

3.3.2 Primary components 73

3.3.2.1 Reservoirs and nitrogen supply 73

3.3.2.2 Pumps 74 3.3.3.3 Oven 74 3.3.2.4 Feed tubes 74 3.3.2.5 Microreactor 74 3.3.3 Secondary components 76 3.3.3 1 Micromixers 76

3.3.3 2 Back pressure regulator (BPR) 77

3.3.4 Auxiliary components 78

3.3.5 General procedure of microreaction 79

3.3.6 High pressure reaction 81

3.3.7 Reaction under high shear rate 82

3.4 Characterizations

83

3.4.1 Precipitation of polymer 83

3.4.2 1H NMR analysis 83

3.4.2.1 Conversion calculation during homopolymerization 84

3.4.2.2 Conversion calculation during copolymerization of DMAEMA & BzMA 84

3.4.2.3 Conversion calculation during branched polymerization 85

3.4.3 Gel permeation chromatography 87

3.4.3.1 Determination of steady state of continuous polymerization microreactor.88

3.4.3.2 Branching efficiency 89

3.4.4 Rheological behavior of polymerizing solution 90

3.4.4.1 General procedure 90

3.4.4.1 Intrinsic viscosity and different rheological parameters 90

3.4.5 Pressure drop determination 91

(21)

4. EFFECT OF PREMIXING AND OPERATING PARAMETERS ON

REACTION RATE

Preface 94

4.1 Intensifying the ATRP synthesis of statistical copolymers by continuous

micromixing flow techniques

4.1.1 Introduction 96

4.1.2 Materials and methods 97

4.1.3 Results and discussion 97

4.1.4 Summary 103

4.1.5 Supporting Information 104

References

106

4.2 Atom Transfer Radical Polymerization in continuous-microflow: effect

of process parameters

4.2.3.1 Effect of temperature on polymerization in microreactor 110

4.2.3.2 Effect of pressure 111

4.2.3.3 Effect of shear rate on polymerization 114

4.2.4 Summary 117

References

123

5. COIL FLOW INVERSION AS A ROUTE TO CONTROL

POLYMERIZATION IN MICROREACTORS

Preface 127

5.1 Effect of coil flow inversion on macromolecular characteristics

5.1.3 Results and discussions 130

5.1.3.1 Linear polymer synthesis 130

5.1.3.2 Branched polymer synthesis 131

5.1.4 Summary 137

(22)

5.2. Flow inversion: an effective mean to scale-up controlled radical

polymerization in tubular microreactors

5.2.3.1 Effect of reactor geometry 146

5.2.3.2 Effect of reactor diameter 147

5.2.3.3 Effect of reactor length 149

5.2.3.4 Process parameters 150

5.2.4 Summary 152

References

158

(23)

(24)

Chapter 1. Introduction

2

Introduction

Nowadays polymers and polymeric materials are used for unconventional applications beacause of their functional properties (e.g. electronics, solar cells, biomedical …) conversely to conventional applications which rely on bulk properties (e.g. automotive parts, textile …). Such high end applications demand for well controlled characteristics of macromolecules. Minor changes in these characteristics can influence the final properties.1-3 Poly(2– (dimethylamino)ethyl methacrylate) (PDMAEMA) is one of such functional polymers. pH and thermo responsive nature of PDMAEMA makes it a preferred candidate for a variety of demanding applications like drug and non-viral gene deliveries among others.4-11 To extract maximum benefit of such polymer, high control on its macromolecular characteristics is a prerequisite.

Controlled radical polymerization (CRP) techniques are among the most effective chemical-based methods to control macromolecular characteristics of a polymer; unlike the established free radical polymerization technique which suffers from termination reactions resulting in a broad molecular weight distribution. Among all the CRP techniques, Atom Transfer Radical Polymerization (ATRP) is the most widely used technique from an academic and industrial research point of view.12 However due to its inherent kinetic scheme, which exhibits an equilibrium reaction between dormant and propagating species, ATRP suffers from low productivity resulting from a slow kinetics.

On the other hand, any polymerization reaction is affected by the process conditions especially in case of concentrated solutions. Besides, polymers are known to be process products. Therefore the process conditions should be well adapted to avoid any detrimental effect on the chemical control of the macromolecular characteristics. New process considerations referred as microreaction technology has emerged during the last decade which can at the same time intensify polymerization processes while maintaining or improving the control over polymer characteristics.

In this context, the PhD work aims at intensifying ATRP processes for the production of DMAEMA-based polymers by relying on microreaction technology tools (microreactor, micromixers) and process parameters (reactor geometry, temperature, pressure…).

(25)

3 in microfluidic systems is relatively new and lot of benefits can be achieved from characteristics features of microdevices.

Chapter 3 details materials and methods used during this work. Among other, the polymerization system composition and procedures followed during experiments in batch reactor will be presented. Microreaction setup used for polymer continuous-flow synthesis will be illustrated in details along with operating parameters. Microreactors with different dimensions and geometries will be described. Procedures for high pressure and high shear reaction will be explained. Finally the characterization section will highlight all the different methods and techniques used to characterize reactors and determine macromolecular characteristics.

Chapter 4 will present the effect of different micromixing principles on the characteristics of statistical DMAEMA-based copolymer. Bilaminaion (T-Junction), impact jet (KM Mixer) and interdigital multilamination (HPIMM) micromixers were considered for the ATRP copolymerization of 2-(dimethylamino)ethyl methacrylate (DMAEMA) with benzyl methacrylate (BzMA). Polymers of two different compositions of BzMA (20 mol.% and 40 mol.% BzMA) were synthesized in coiled tubular (CT) microreactor. In the second section of this chapter, ATRP polymerization of DMAEMA in CT reactors under different operating conditions (temperature, pressure and shear rate) is discussed. It gives a clear idea how such polymerization reaction can be accelerated. This section also highlights the effect of CT reactors diameter.

(26)

4 Finally chapter 6 will highlight the overall outcome of this work. Experience and knowledge gained during this thesis should be quite helpful for future work. Thus suggestions and recommendations are to be given to proceed further.

This PhD work is part of a larger project named DIP², funded by the French Research Agency (ANR grant n ° 09-CP2D-DIP²), which aim was to intensify a CRP process for the production of architecture-controlled polymers. It comprised an experimental work (this thesis) as well as a numerical work for the geometry optimization of CFI reactors.

References

(1) Ramkissoon-Ganorkar, C.; Liu, F.; Baudys, M.; Kim, S. W. Journal of biomaterials

science. Polymer edition 1999, 10, 1149.

(2) Rathfon, J. M.; Tew, G. N. Polymer 2008, 49, 1761.

(3) Xu Zheng, Z. T.; Xie, X.; Zeng, F. Polymer Journal 1998, 30, 5.

(4) Mao, J.; Ji, X.; Bo, S. Macromolecular Chemistry and Physics 2011, 212, 744.

(5) Pietrasik, J.; Sumerlin, B. S.; Lee, R. Y.; Matyjaszewski, K. Macromolecular

Chemistry and Physics 2007, 208, 30.

(6) Yuan, W.; Guo, W.; Zou, H.; Ren, J. Polymer Chemistry 2013, 4, 3934.

(7) Jiang, X.; Lok, M. C.; Hennink, W. E. Bioconjugate Chemistry 2007, 18, 2077.

(8) Lin, S.; Du, F.; Wang, Y.; Ji, S.; Liang, D.; Yu, L.; Li, Z. Biomacromolecules 2007, 9, 109.

(9) Park, T. G.; Jeong, J. H.; Kim, S. W. Advanced Drug Delivery Reviews 2006, 58, 467. (10) Yancheva, E.; Paneva, D.; Maximova, V.; Mespouille, L.; Dubois, P.; Manolova, N.;

Rashkov, I. Biomacromolecules 2007, 8, 976.

(27)

Chapter 2. Background litterature

5

CHAPTER 2 BACKGROUND LITERATURE

Preface 6

2.1 Aspects of Controlled Radical Polymerization (CRP)

7

2.1.1 General overview 7

2.1.2 Special considerations for ATRP 8

2.1.3 Intensification of CRP in batch reactors 9

2.1.3.1. Acceleration of NMP 9 2.1.3.2. Acceleration of RAFT 10 2.1.2.3 Acceleration of ATRP 12 2.1.4 Summary 14

2.2 Microreaction technology

14 2.2.1 Microreactors 16 2.2.1.1 Different types 16 2.2.1.2 Special features 19

2.2.2 Microdevices for polymer synthesis 23

2.2.2.1 Microreactor setup 23

2.2.2.2 Micromixers 25

2.2.3 Benefits of microdevices for polymers and copolymers synthesis 31

2.2.3.1 Polymers with controlled macromolecular characteristics 32

2.2.3.2 Polymers with controlled chemical composition 40

2.2.3.3 Polymers with controlled architecture 45

2.2.3.4 New operating windows 49

2.2.4 Online monitoring 55

2.2.5 Summary 55

2.3 Conclusion

56

(28)

6

Preface

Controlled radical polymerizations (CRP) find wide acceptance due to their efficient control over polymerization and macromolecular characteristics. However, slow rate of reaction is a major challenge for CRP. Acceleration of CRP without losing control over polymerization will be beneficial from industrial perspective. Several methods to achieve faster kinetics of CRP in batch reactor are reported in literature. Overview of such strategies to accelerate different CRP like NMP, RAFT and ATRP are briefly discussed in the first part of this chapter.

On the other hand, microreaction technology is by now largely considered for process intensification in fine chemical and pharmaceutical synthesis. In the field of polymer reaction engineering, microreaction is just a decade old and genuine fear of increased viscosity and low throughput are few hurdles which force researchers and industries to undermine its potential. Nevertheless some applications of microfluidics can be found in the field of polymer synthesis but quite few concern the rate enhancement of CRP. To have a clear understanding about this budding technique and its impact in the field of polymer reaction engineering a detailed discussion is needed. Therefore polymerization micro-chemical plants for the synthesis of polymers and copolymers composed of microdevices are described and commented in the second part of this chapter. Due to their unique characteristics, microdevices allow rapid heat removal and mixing. This contributes to improve the control over the polymerization by reducing or eliminating mass transfer limitations and hot spot formation. This chapter also highlights how fast mixing and heat management allow obtaining macromolecules with better controlled characteristics (molecular weights and narrower molecular weight distributions), compositions and architectures.

This chapter is partially adapted from the following online review:

- C. A. Serra, D. Parida, F. Bally, D.K. Garg, Y. Hoarau, and V. Hessel, "Micro-Chemical

Plants" in «Encyclopedia of Polymer Science and Technology »; Wiley-VCH, Weinheim

(29)

7

2.1. Aspects of Controlled Radical Polymerization (CRP)

2.1.1 General overview

Simple reaction conditions, faster kinetics and ability to polymerize a wide range of vinyl monomers makes Free Radical Polymerization (FRP) a robust and economical process to synthesize commodity polymers.1 However, unavoidable termination and transfer reactions due to high reactivity of transient species result in broad molecular weight distributions. Poor control over molecular weight distribution, composition and architecture limits application of FRP in the field where above mentioned polymer characteristics are of prime importance. To overcome limitations of FRP some techniques were developed in last few decades and known as controlled radical polymerization (CRP). CRP allows synthesizing polymer with controlled characteristics and architecture for high end applications. Among all the CRP techniques developed Nitroxide Mediated Polymerization (NMP),4-6 Reversible Addition-Fragmentation chain Transfer polymerization (RAFT)7-9 and Atom Transfer Radical Polymerization (ATRP)10,11 are most widely used and reported.

Scheme 2.1. General scheme of NMP, RAFT and ATRP.12

(30)

8 deactivating species (X) as shown in scheme 1. These new dormant species are stable and can be reactivated easily with the help of heat, light or catalyst. After reactivation, the newly formed radical can propagate like in normal FRP. In case of radical–radical termination, concentration of deactivating species (X) increases in the system. Detailed mechanisms of different CRP techniques are extensively discussed in the literature.2,11-13

Controlled characteristics of CRP enable to synthesize gradient, block, star copolymers and polymers with controlled graft densities.14-16 From adaptability and tolerance point of view CRP systems are superior to conventional FRP. They are more tolerant towards different solvent systems17-20 and functionality of monomers.21-23 These features of CRP draw lot of attention not only from academia, but also from industries.

2.1.2. Special considerations for ATRP

ATRP was devised by two independent groups in mid 90s.8,9 Since then many modifications and developments have taken place and by the years ATRP is emerging as a powerful technique for the synthesis of polymers with controlled molecular weight, low polydispersity, controlled chain end functionality, morphology and composition. The ATRP method relies on the reversible homolytic cleavage of an alkyl halide initiator molecule in the presence of transition metal salt complexed with a suitable bi- or tridentate ligand acting together as catalyst. After cleavage the monomer starts to be incorporated in growing chains.

Considering the wide acceptance and its relevance to the thesis, ATRP is discussed in brief.12 The mechanism involved to control the polymerization is the atom transfer between growing chains and catalyst. Like other CRP, dynamic equilibrium of ATRP is of prime importance to get a controlled polymer characteristics.

Scheme 2.2. Reactions responsible for equilibrium in ATRP, where RX represents alkyl

(31)

9 Equilibrium in ATRP (KATRP) is mainly governed by 4 set of reversible reactions shown in scheme 2.2.24 These reactions are (i) homolysis of alkyl halide bond (KBH), (ii) oxidation of metal complex (indicated as KET, the equilibrium constant for electron transfer), (iii) reduction of halogen to halide ions (KEA, equilibrium constant for electron affinity) and (iv) association of halide ion with metal complex (KX).

ATRP allows producing functional polymers quite easily. Two methods are commonly used. Post functionalization method relies on the modification of the chain end. Indeed carbon-halogen chain end present in an active macromolecule can form hydroxyl, allyl, azido and ammonium or phosphonium end groups by nucleophilic substitution.25-27 Preparation of polymer having amino end groups is also reported in literature. It is achieved by the substitution of the halogen end groups with azides followed by the conversion of the azide groups to phosphoranimine end groups and finally hydrolyzed to form amino groups.24-27 The second method of preparing polymer having different functionalities employs initiator with functional groups.

2.1.3. Intensification of CRP in batch reactors

Slow reaction rate of CRP, originating from the equilibrium reaction, is one of the major challenges on its way towards wide acceptability in commercial applications. To extract maximum benefits of CRP, polymerization rate needs to be enhanced without sacrificing its controlled nature. Acceleration can be achieved by changing chemical system and by process conditions. Such methods to accelerate CRP are discussed in brief in the following section.

2.1.3.1. Acceleration of NMP

(32)

10 power mode and dynamic mode.28 They observed a significant acceleration in NMP rate in pulsed power mode compared to dynamic mode.

Figure 2.1. Kinetic plots of ln([M]0/[M]) vs. reaction time for bulk polymerization of styrene

under microwave (100 and 200 W) and conventional heating (CH) at 125 ºC.30

Accelerator molecules like photo-acid generators were also used to enhance the rate of NMP reaction. One such example is the use of (4-tert-butylphenyl) diphenylsulfonium triflate (tBuS) as an accelerator molecule during Nitroxide Mediated photo polymerization of methacrylic acid. This photo-NMP was carried out using azoinitiators and 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO) as the mediator. Larger accelerator/mediator ratios resulted in increased polymerization rates both in bulk and solution.31 Molecular weight distribution was found to decrease with increase in accelerator/mediator ratio and solvent content in the system.

2.1.3.2. Acceleration of RAFT

(33)

11

Figure 2.2. Relative conversion of monomers with time in different heating systems.33

Pressure is another parameter which was found to have a significant impact on polymerization kinetics of RAFT polymerization.38-40 Initially high pressure RAFT polymerization was used to polymerize methyl ethacrylate which is a monomer difficult to polymerize at ambient pressure because of steric hindrance of α-ethyl substituent.38 Another example was reported by Arita et al. who observed a rapid increase in styrene polymerization rate (by a factor of 3) going from ambient pressure to 2500 bars as shown in Figure 2.3.39 Interestingly reduced molecular weight distribution was observed in case of high pressure polymerization indicating better control over polymerization.

Figure 2.3. Effect of pressure on rate of polymerization of styrene at different cumyl

(34)

12 2.1.2.3 Acceleration of ATRP

Like other two CRP techniques discussed above, there are also different methods reported to accelerate ATRP. One such method relies on use of alcohol/water (95/5 vol. %) as solvent and mixed transition metal catalyst (Fe(0)/CuBr2 = 1/0.1). It was found that within few hours (16 h

instead of days) 98 % conversion was reached at 30 ºC polymerization temperature. This method has the advantage of cheaper solvent and significant reduction of toxic copper from the polymerization system.40

Another way to accelerate ATRP in case of silica nanoparticles surface initiated process was reported by Liu et al.41 Usually surface initiated ATRP (SI-ATRP) polymerization rate is much slower than volume polymerization rate due to geometric constraints of high density polymer chains on the surface. High surface density of chains leads to irreversible termination and slows down the polymerization rate. To reduce termination, Liu et al. used self polymerized polystyrene chains during SI-ATRP of styrene. They observed acceleration of SI-ATRP as a faster kinetics was obtained with increase in polystyrene content (Figure 2.4). However, such strategy had no effect on polymerization rate during volume polymerization of styrene.

Figure 2.4. SI-ATRP of styrene on silica nano particles using 0.5% polystyrene (a) and 1%

polystyrene in polymerization system.41

Unlike two other CRP techniques (NMP and RAFT) high pressure route to accelerate ATRP was more explored.42-47 High pressure enhances propagation (higher kp) and reduces

termination (lower kt) rates. As a result polymerization rate increases significantly and

(35)

13 methyl methacrylate under high pressure (up to 2500 bars) in acetonitrile. Equilibrium constant (KATRP) at ambient pressure was 3.8 × 10−6 and increases by one order of magnitude

with an increase in the pressure up to 2500 bars (Figure 2.5).46 Propagation kinetic constant (kp) also increased by nearly one order of magnitude with pressure. Overall consequence was

an increase in polymerization rate by two orders of magnitude from atmospheric pressure to 2500 bars.

Figure 2.5. Pressure dependence of KATRP for ATRP of MMA at 25 ºC.46

High pressure polymerization also allows carrying out polymerization at elevated temperature without increasing termination.49 It accelerates the polymerization further without affecting the characteristics of ATRP as shown in Figure 2.6. Interestingly it was found that very low level of Cu catalyst (~ 100 ppm) is required in case of high pressures.

Figure 2.6. Plot of ln([M0]/[M]) vs time for ATRP of n-butyl acrylate at different

(36)

14

2.1.4 Summary

Two distinctive strategies can be worked out in batch reactors to accelerate CRP, namely chemical- and process-based methods. The former is likely to be strongly dependent of the polymerization system like in the above mentioned SI-ATRP and thus should be adapted for each new case. The latter is composed of microwave- and pressure-assisted polymerization processes. However failure to accelerate polymerization of non-polar monomers like styrene is a serious limitation of the microwave method.34 High pressure was found to be beneficial in all CRP techniques from kinetics point of view. However polymerization at such high pressure (few thousands bars) is always troublesome and requires specialized equipments. This makes the process very expensive and unsafe. Nevertheless for these two methods, controlled nature of the CRP was not altered.

On the other hand, there is a new technique available which has the ability to accelerate polymerization rate and improve the control over macromolecular characteristics. This technique refers to microreaction technology. Features, abilities and advantages of this new technique in the field of polymer synthesis are discussed in the following section.

2.2 Microreaction technology

(37)

15 which are greatly affected by the process parameters which in turn influence their properties (e.g. mechanical, thermal, and processing). Viscosity increase is considered as one of the foremost important parameters.57 This increase, which can reach up to 7 decades for bulk and highly concentrated solution processes, is followed by a respective decrease of the reactants molecular diffusion coefficients and the reduction of mass and thermal transports.57,58 As an overall consequence, the polymerization rate may locally increase. This is the gel or Trommsdorff effect. This uncontrolled acceleration of the polymerization reaction rate leads to the broadening of the chain lengths distribution and the polydispersity (PDI) increases. Another frequent problem encountered with polymerization processes is the removal of the heat released by the reaction. For exothermic polymerization reactions, insufficient thermal transport towards the cooling system (e.g. double jacket or immerged serpentine) may lead to thermal runaways. The consequences on polymer properties are usually twofold. First of all, a significant broadening of the chain lengths distribution can be observed. Secondly, beside non-reproducible results, the number-average chain length may significantly vary (several orders of magnitude) from the expected value. Mixing is also an extremely important issue in homogeneous polymerization processes. Low quality mixing between on one hand, the monomer phase, and on the other hand, the solvent and initiator phase, will generate large local concentration gradients, often termed segregation.55 Since the propagation reaction is usually very fast zones of high monomer concentrations will produce high molecular weight polymers and release a lot of energy. Thus, a hot spot is formed which can lead to degradation of the polymer and/or the monomer.

(38)

16

2.2.1 Microreactors

2.2.1.1 Different types

A microreactor (or several) constitutes the heart of any micro-chemical plant and its selection depends on the type of reaction to be performed and operating conditions (e.g. temperature, pressure). Microreactors can be broadly classified into three different categories as shown in scheme 2.3.

Scheme 2.3. Different types of microreactors and their areas of application.

2.2.1.1.1 Singular type

(39)

17

Figure 2.7. Image of a singular type microreactor.81

2.2.1.1.2 Capillary type

Capillary type microreactors are tubes of internal diameter less than one millimeter and length ranging from millimeters to few meters e.g., those commonly used for HPLC applications. Poly(tetrafluoroethylene) (PTFE), poly(ether ether ketone) (PEEK) and stainless steel are the most common materials for this type of microreactors allowing to carry out chemical reactions under a wide range of pressures from few bars up to several hundreds. This kind of microreactors do not have inbuilt mixers as in case of the singular type and microstructured reactors. The idea is that premixing can be done without initiating reaction and rising (substantially) temperature sets the reaction starting point. This indeed works out well for many known homogeneous and heterogeneous liquid-liquid reactions. Microfluidic reactors made from tubing are relatively cheap and long reactor lengths can be easily achieved. This enables chemical reactions to be conducted with few seconds of residence time to several hours to achieve high conversion; yet in the latter case it is obtained at expense of reducing the flow rate. The throughput of capillary microreactors is higher as compared to singular type microreactors mainly to a higher internal volume. Thus, flow rates of up to hundreds of milliliters per hour are commonly employed. However, careful selection of the reactor material should be made depending upon the reaction conditions like pressure, oxygen sensitivity or resistance to solvent etc. For example polymeric capillaries are not suitable for high pressure reactions. Depending on the thickness of the wall of capillary, oxygen can diffuse through it. It can be detrimental to oxygen sensitive reactions like RAFT polymerization.80

(40)

18 ozone are easily fed into a liquid medium in this way. Naturally, a (released) gas can be immediately removed in the same manner.83

2.2.1.1.3 Microstructured

Complex geometries on microscale can be realised by this type of microreactors through flexible interconnection and integration of highly specialised mixers, reactors, and other process functions along a flow process line. These geometries are designed to provide additional functions (e.g. mixing, separation, delay loop) to the microdevice beside the locus of a chemical reaction. Thus, these microreactors can be equipped with upstream micromixing zones of even embedded mixing elements in the reaction microchannels. Figure 2.8a shows an example of a microstructured reactor, named cyclone mixer, designed for the production of foams and gas-liquid dispersions in general. It has 11 microstructured stacks each containing 3 groups of nozzles for supplying reagents. The gas and liquid injection nozzles are 30 µm and 50 µm wide respectively. Inside the reaction area, the gas bubbles form a cyclone-like pattern within the liquid which deform and coalesce in smaller microchannels. This pattern is very advantageous when efficient contact between a slurry catalyst and a gas phase is necessary in liquid-gas reactions. The mixer can also perform liquid-liquid reactions efficiently.84 Microstructured reactors as shown in Figure 2.8b were designed for gas-liquid reactions and the microstructures were designed to acts like a catalyst retainer.85 Numerous examples of such microreactors have reported in the literature performing different functions like catalytic reaction or separation.77,86-88 These microstructured reactors are efficient in handling most kind of reactions. However they are not preferred for polymerization reactions because of their inability to handle viscous solutions. Moreover, this type of microreactors is expensive and may require long fabrication time.

(41)

19 (b)

Figure 2.8. (a) Glass cyclone mixer78, (b) a microstructured reactor (1) bifurcating the reactant to 64 inlets which pass over a catalyst bed, (2) picture showing microreactor packed

with 60 µm glass beads, (3) SEM image of the microreactor. 85 2.2.1.2 Special features

2.2.1.2.1 Efficient thermal management

(42)

20

Figure 2.9. Typical temperature profile for different reactors and formation of by products.89

Figure 2.10. Surface to volume ratio of different microdevices and conventional reactors.90

2.2.1.2.2 Enhanced mixing

(43)

21 and semi diluted polymerization reactors) and design considerations impose limitation on turbulent mixing at macroscale. Unlike macroreactors, mixing in microreactors is only governed by diffusion. Typical mixing times achieved are usually in the range of few milliseconds; even by order of magnitude faster is doable. Therefore, microreactors are suitable for mass-limited reactions which are often encountered in the synthesis of polymers due to the viscosity increase during the course of the reaction.

2.2.1.2.3 Residence time control

(44)

22

Figure 2.11. Principle for generation and reaction of unstable short-lived reactive

intermediates based on residence time control in a flow microreactor.91 2.2.1.2.4 Conformation of macromolecules

Extensional flow field generated in micro dimension due to laminar flow can make macromolecules aligned. Yamashita and coworkers have observed elongated conformation of DNA strands along the direction of flow within a microfluidic setup as shown by the optical micrograph in Figure 2.12b compared to a coiled conformation observed during non flowing condition as shown in Figure 2.12a.92 It is believed that aligned conformation of macromolecules makes active sites more accessible for reaction by reducing the steric hindrance. However, this feature was not well explored even if a possible application may involve the synthesis of functional polymers with dendritic structure.

Figure 2.12. Optical micrographs showing coiled DNA strands at rest (a), stretched DNA

strands due to flow in microchannels (b).92

(a)

(45)

23 2.2.1.2.5 Modularity and versatility

This feature concerns the possible arrangement of multiple microreactor units in series or in parallel by standardized interconnects, placements, or dimensions. This might be quite helpful for combinatorial synthesis approach or simply for throughput increase. Indeed, for the latter case, it enables to retain the same characteristics that are achieved on a single unit without need for up scaling, this is known at the numbering-up approach. Moreover arrangement in series gives the flexibility of multi-stage operations allowing to easily varying the product quantity and diversification. Therefore, the modularity and versatility of microreactors can be advantageously used for the production of libraries of copolymers with different compositions.

2.2.2 Microdevices for polymer synthesis

2.2.2.1 Microreactor setup

(46)

24 and temperature probes can also be classified as secondary components. Back pressure regulators are used at the exit of the microreactor to control the pressure during polymerization and maintain uniform reaction condition. They also allow conducting the reaction above the boiling point of reactants as the pressure inside the reactor can be maintained above atmospheric pressure. Auxiliary components like two-way valves, three-way valves are quite helpful to change the composition of reactions, remove waste during cleaning of the microdevice. Sometimes non invasive online monitoring systems (optic or spectroscopic methods) can be used to monitor the reaction and this kind of equipments falls in the category of auxiliary components. It is worth mentioning that a microreactor setup is a modular device. Thus another microreactor can be attached downstream to the first one such as the inlet stream of the first microreactor serves as one feed line of the second. This is quite helpful during synthesis of block copolymers for instance.

Figure 2.13. A typical microreaction setup for polymerization showing: Reservoirs (a), HPLC

Pumps (b), Micromixer (c), Pressure sensor (d), Oven (e), Tubular microreactor (f), Back pressure regulator (g), Three-way valve (h).

(47)

25 concentrations of reactants (stoichiometry), polymer libraries, with polymers presenting controlled molecular masses, could be rapidly generated.

(a)

(b)

Figure 2.14. A microchannel-based reactor used for ATRP (a), SEC traces of poly(HPMA)

produced from different flow rates and monomer to initiator concentration ratios (b).93 Block copolymers of poly(ethylene oxide-2-hydroxypropyl methacrylate) (PEO-b-PHPMA) have also been synthesized by Wu et al., using the same strategy.94 A three-input-one-output chip was used for the mixing area and two other one-input-one-output microchannels were connected to increase the reactor length. A macromolecular PEO initiated the polymerization and a wide range of well-defined second blocks have been obtained by varying the total flow rate; similar material variations were achieved by changing polymerization time or initiator concentration.

2.2.2.2 Micromixers

(48)

26 numbers are thus usually quite low and the flow remains laminar or in the transition regime when the flow velocity is set high (Re can then be a few times 100). In the true laminar regime, the two fluids entering a microdevice flow parallel to each other and mixing takes place at the interface by mass diffusion. If the Re number is higher and even in the intermediate regime, a so-called intertwining regime sets in with strong mutual exchange of fluid segments via recirculation, i.e. the mixing mechanism becomes convective. This is quite difficult to achieve, however, for viscous flows such as polymeric solutions. To speed up the pure diffusional mixing, the length of diffusion must be as short as possible. This was the starting point in the design of different types of micromixers. They can operate either by multilamination of the fluids streams to be mixed into lamellae of low thicknesses (few tens of microns), by splitting and recombining the main flow or by impacting jet streams at high velocity on a spot of small dimension (less than one square millimeter).

Figure 2.15. Relation between gravitational, inertial, viscous forces to interfacial forces as a

function of channel size and velocity.95

(49)

27 Being very reactive, acrylic monomers react fast and poor mixing of monomer and initiator solutions leads to uneven propagation of polymerization. Thus very high molecular weight compounds are generated which can make them at some point insoluble in the mixture or solvent and unreacted monomer leading to their precipitation. To avoid such fouling, a multilamination type micromixer (IMM, Mainz, Germany) was used prior to the static mixer. This micromixer laminates both inlet solutions into 36 lamellae of 25 µm thickness. The generation of such thin lamellae and the subsequent rapid and efficient mixing prevented the precipitation of the polymer as seen in Figure 2.16b. Thus the polymer synthesized had a number-average molecular weight of 104 g/mol.

(a) (b)

Figure 2.16. Fouled static mixer due to precipitation of high molecular weight acrylic

(50)

28 2.2.2.2.1 Multilamination

As the name implies, in multilamination micromixers incoming fluid streams are divided into multiple lamellae as shown in Figure 2.17a.104,105 Then the lamellae are alternatively arranged by means of a specific designed microstructure (Figure 2.17a). This process brings down the mixing time from tenths to few seconds by enhancing diffusion. If combined with flow compressing (geometric focussing) millisecond mixing can be achieved. The High Pressure Interdigital Multilamination Micromixer (HPIMM, IMM, Mainz, Germany) is suitable for operation up to 200 bars and suits therefore well to the needs of some polymerization reactions (Figure 2.18a). Mixing efficiency depends upon the number and width of mirochannels present in the laminating element known as inlay (Figure 2.18b) which is housed in the lamination section, and on the focusing ratio in the so-called slit located in the top micromixer part and on top of the inlay.

(a) (b)

Figure 2.17. Comparison of multilamination and flow focusing adapted to multilamination.

Unfocused and focused (a), Focused with large focussing ratio–SuperFocus micromixer (b).105,106

(a) (b)

(51)

29 A remarkable improvement for this kind of micromixers was made possible by systematic variation and investigation of the so-called flow-focusing technology (Figure 2.17a).106This technique allows generating lamellae of nearly 1 µm thickness by forcing them to flow through a converging chamber. In this way, thinner lamellae can be achieved which is difficult to obtain by simple microstructuring. Generation of such thin lamellae allows reducing the mixing time down to few milliseconds, as said above. Such modifications not only enhance the mixing but also increase the throughput up to 8-10 L/h (since the large entry cross-section of the focussing chamber needs to be fed by a very large number of parallel lamellae/feeds).106

2.2.2.2.2 Split and recombine

This kind of micromixers relies on the splitting of the incoming streams into multiple substreams and thereafter on their recombination.99 This operation of splitting and recombination can be divided into different steps as shown in Figure 2.19. Starting from splitting of multilayered streams perpendicular to the lamellae orientation into substreams, followed by the re-alignment of substream lamellae and finally by the recombination of these lamellae. It should be kept in mind, however, that the splitting in Figure 2.19 is ideal and may be – more or less – different from reality. Even under laminar conditions, it is known that lamellae deform due to shear forces, giving rise, e.g. to convex and concave-shaped lamellae with varying local thickness (at one cross-section) and along the flow axis varying average thickness. Moreover, for flows at higher Re number, the insetting recirculation finally will disrupt the lamellae causing actually finer dispersed cross-sections as given in Figure 2.19, which means much improved mixing as compared to simple serial lamination.

(52)

30

Figure 2.19. Schematic of split and recombine principle.99

(a)

(b)

Figure 2.20. Schematic view and cross-sectional view of the SAR micromixer (a),

(53)

31 2.2.2.2.3 Impact jet

The working principle of such kind of micromixers is based on both providing high kinetic energy which after jet collision enables turbulence in a much confined mixing chamber (‘atomization’). Inlet fluid streams are first divided into several multiple micro jets by using nozzles; then these jets are allowed to collide at the mixing chamber and finally guided to a microchannel.72,109 Micromixers developed by Synthese Chemie (Lebach, Germany) and KM mixer (Fujifilm Corporation, Kyoto, Japan) are based on this principle (Figures 2.21 & 2.22). Besides for mixing leading to a mixed single-phase such as given in homogeneous polymerisations, these micromixers can be advantageously used for dispersion making essential for heterogeneous (emulsion) polymerisations and even for particle making.72

Figure 2.21. Micromixer developed by Synthesechemie (Lebach, Germany).72

Figure 2.22. Schematic drawing of the flow inside a KM mixer.109

2.2.3 Benefits of microdevices for polymers and copolymers synthesis

(54)

32 constraint, which makes them suitable candidates for microreactors. Some beneficial features of microreactors compared to macrodevices are discussed in the following sections.

2.2.3.1 Polymers with controlled macromolecular characteristics

Free Radical Polymerization (FRP) is a polymerization technique that enables production of polymers on an industrial scale due to favourable operating conditions and reaction time. However, the polydispersity of the final product is often high because of poor control over the polymerization course. This is mainly due to inefficient temperature control within the whole volume of batch reactor, leading to undesired hot spots. In order to improve the heat transfer, microreactors with a surface to volume ratio much larger than conventional heat exchangers have been developed. The AIBN (2,2-Azobis(isobutyronitrile)) initiated FRP of 5 different monomers was investigated by Iwasaki et al.110 The micro-chemical plant (Figure 2.23) was composed of a T-shape micromixer (M1: 800 µ m I.D.), a primary microtube for achieving complete mixing (R1 250 µm I.D., 2 m length), a microtubular reactor immersed in a 100°C oil bath (R2: 500 µm I.D., 9 m length) and a microtube immerged in a water bath at 0°C for polymerization quenching (R3: 500 µm I.D., 1 m length). The results were compared with those obtained in a conventional macroscale batch reactor. For butyl acrylate (BA), the molecular weight distribution was found much narrower than for the batch reactor as seen in Figure 2.24. The difference was smaller but still noticeable for benzyl methacrylate (BMA) and methyl methacrylate (MMA) and almost null for vinyl benzoate (VBz) and styrene (St). Authors claimed that the observed results are directly related to the superior heat transfer ability of the microtubular reactor. The more exothermic the polymerization reaction, the more effective is the microdevice to control the molecular weight distribution. Similar results conducted on the polymerization of styrene were also obtained by Leveson et al.111with a microtubular reactor of 500 µm I.D. and a tubular reactor of 4.2 mm I.D.

(55)

33

Figure 2.24. Molecular weight distribution of poly(butyl acrylate) produced in the

microreactor (plain line) and in the macroscale batch reactor (dashed line). Residence time was 4 min.110

(56)

34 reactors of 1950 mm length with successively increasing internal diameter: 250, 500 and 1000 µm. Finally, the fifth section allowed merging all 8 microtubular reactors into one microtube at low temperature to quench the polymerization. AIBN-initiated FRP of BA performed in a single tube of varied inner diameters (250 µm, + 500 µm, + 1000 µm) gave similar results to those obtained with this Type-2 numbering-up reactor demonstrating that a good flow uniformity was achieved in this latter system. Finally, as depicted in Figure 2.25, a pilot plant was constructed based on the Type-2 numbering-up reactor. This pilot has been operated continuously for 6 days producing up to 4 kg of PMMA without any increase in the pressure or reactor temperature. The quality of the polymer was constant over one week of operation as seen in Figure 2.26. This pilot plant operation demonstrates that microdevices can be applied to production of polymers at comparatively large scale.

Figure 2.25. Photograph of the microchemical pilot plant.112

Figure 2.26. Variation of the number-average molecular weight and polydispersity index of

poly(MMA) against days of operation.112